# Micro-Batch Training with Batch-Channel Normalization and Weight Standardization

arXiv: https://arxiv.org/abs/1903.10520 · HTML: https://mlsanity.com/paper/1903.10520/

## Key facts

- Cited by 46 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 25 Mar 2019.

## Built on by (most cited first, 30 of 46)

- [High-Performance Large-Scale Image Recognition Without Normalization](https://mlsanity.com/paper/2102.06171/) (2021)
- [Large-Scale Representation Learning on Graphs via Bootstrapping](https://mlsanity.com/paper/2102.06514/) (2021)
- [Realistic One-shot Mesh-based Head Avatars](https://mlsanity.com/paper/2206.08343/) (2022)
- [Characterizing signal propagation to close the performance gap in unnormalized ResNets](https://mlsanity.com/paper/2101.08692/) (2021)
- [Graph Barlow Twins: A self-supervised representation learning framework for graphs](https://mlsanity.com/paper/2106.02466/) (2021)
- [The Evolution of Out-of-Distribution Robustness Throughout Fine-Tuning](https://mlsanity.com/paper/2106.15831/) (2021)
- [Does Your Dermatology Classifier Know What It Doesn't Know? Detecting the Long-Tail of Unseen Conditions](https://mlsanity.com/paper/2104.03829/) (2021)
- [Rotational Equilibrium: How Weight Decay Balances Learning Across Neural Networks](https://mlsanity.com/paper/2305.17212/) (2023)
- [Online Training Through Time for Spiking Neural Networks](https://mlsanity.com/paper/2210.04195/) (2022)
- [MQBench: Towards Reproducible and Deployable Model Quantization Benchmark](https://mlsanity.com/paper/2111.03759/) (2021)
- [Supervised Transfer Learning at Scale for Medical Imaging](https://mlsanity.com/paper/2101.05913/) (2021)
- [Contrastive Representation Learning for Hand Shape Estimation](https://mlsanity.com/paper/2106.04324/) (2021)
- [Does enhanced shape bias improve neural network robustness to common corruptions?](https://mlsanity.com/paper/2104.09789/) (2021)
- [Meta Adversarial Training against Universal Patches](https://mlsanity.com/paper/2101.11453/) (2021)
- [Unsupervised Learning of Equivariant Structure from Sequences](https://mlsanity.com/paper/2210.05972/) (2022)
- [Sequential Learning Of Neural Networks for Prequential MDL](https://mlsanity.com/paper/2210.07931/) (2022)
- [SPeCiaL: Self-Supervised Pretraining for Continual Learning](https://mlsanity.com/paper/2106.09065/) (2021)
- [One-Trimap Video Matting](https://mlsanity.com/paper/2207.13353/) (2022)
- [Logit Normalization for Long-tail Object Detection](https://mlsanity.com/paper/2203.17020/) (2022)
- [GC-MVSNet: Multi-View, Multi-Scale, Geometrically-Consistent Multi-View Stereo](https://mlsanity.com/paper/2310.19583/) (2023)
- [Towards Single Stage Weakly Supervised Semantic Segmentation](https://mlsanity.com/paper/2106.10309/) (2021)
- [Revisiting Context Aggregation for Image Matting](https://mlsanity.com/paper/2304.01171/) (2023)
- [Semi-weakly Supervised Contrastive Representation Learning for Retinal Fundus Images](https://mlsanity.com/paper/2108.02122/) (2021)
- [Equivariant Differentially Private Deep Learning: Why DP-SGD Needs Sparser Models](https://mlsanity.com/paper/2301.13104/) (2023)
- [Low-Precision Reinforcement Learning: Running Soft Actor-Critic in Half Precision](https://mlsanity.com/paper/2102.13565/) (2021)
- [Tensor Normalization and Full Distribution Training](https://mlsanity.com/paper/2109.02345/) (2021)
- [Generalization to translation shifts: a study in architectures and augmentations](https://mlsanity.com/paper/2207.02349/) (2022)
- [Federated Training of Dual Encoding Models on Small Non-IID Client Datasets](https://mlsanity.com/paper/2210.00092/) (2022)
- [Fake or Compromised? Making Sense of Malicious Clients in Federated Learning](https://mlsanity.com/paper/2403.06319/) (2024)
- [Weight Rescaling: Effective and Robust Regularization for Deep Neural Networks with Batch Normalization](https://mlsanity.com/paper/2102.03497/) (2021)

## Abstract (arXiv)

Batch Normalization (BN) has become an out-of-box technique to improve deep network training. However, its effectiveness is limited for micro-batch training, i.e., each GPU typically has only 1-2 images for training, which is inevitable for many computer vision tasks, e.g., object detection and semantic segmentation, constrained by memory consumption. To address this issue, we propose Weight Standardization (WS) and Batch-Channel Normalization (BCN) to bring two success factors of BN into micro-batch training: 1) the smoothing effects on the loss landscape and 2) the ability to avoid harmful elimination singularities along the training trajectory. WS standardizes the weights in convolutional layers to smooth the loss landscape by reducing the Lipschitz constants of the loss and the gradients; BCN combines batch and channel normalizations and leverages estimated statistics of the activations in convolutional layers to keep networks away from elimination singularities. We validate WS and BCN on comprehensive computer vision tasks, including image classification, object detection, instance segmentation, video recognition and semantic segmentation. All experimental results consistently show that WS and BCN improve micro-batch training significantly. Moreover, using WS and BCN with micro-batch training is even able to match or outperform the performances of BN with large-batch training.